Stochastic resonance is applied to quantitative analysis for weak chromatographic signal of Sudan I

doi:10.1016/j.physleta.2006.07.012

Physics Letters A

Volume 359, Issue 6, 11 December 2006, Pages 620-623

https://doi.org/10.1016/j.physleta.2006.07.012 Get rights and content

Abstract

Based on the algorithm of stochastic resonance, a new method improved the signal-to-noise ratio (SNR) of weak chromatography signal remarkably. An excellent quantitative relationship can be obtained between concentration and weak chromatography signals of Sudan I which was embedded in the noise background.

Introduction

Noise has always been considered to be a source of disorder and a nuisance to be avoided. In order to avoid the effect of noise and improve the detection limit, chemometric methods, such as fast Fourier transform (FFT), wavelets transform (WT) and various smoothing and filtering algorithms, were widely used. In practice, these methods may result in tiny lost of useful information due to improper truncations. Stochastic resonance (SR), which was first introduced by Benzi and his co-workers to explain the periodicity of Earth's ice age, renders an entirely new way for detecting weak chromatographic signal [1]. SR is a phenomenon that manifests in nonlinear systems, whereby generally feeble input information (such as a weak signal) can be amplified and optimized by the assistance of noise [2]. For a system well characterized by linear-response theory (that is, with linear or “mildly” nonlinear internal dynamics) the signal-to-noise ratio (SNR) at the output must equal the SNR at the input, and any increase in the input noise will result in a decrease in the output SNR. In contrast, the signature of stochastic resonance is an increase in the output SNR with increased input noise [3]. Over the last two decades, SR has continuously attracted considerable attention. Typical SR behaviors have been found in many physical systems [4], [5] and complex systems such as chemical reaction [6], [7] and quantitative structure–activity relationship (QSAR) [8]; however, its application in the analysis field is seldom reported. In this Letter, based on the theory of stochastic resonance, output signal was enhanced by optimizing the parameters a and b, and the detection sensitivity of Sudan I with high performance liquid chromatography with ultraviolet/visible detection (HPLC-UV/VIS) was increased significantly.

Sudan I (1-phenylazo-2-naphthol, CAS 842-07-09) (Fig. 1) is an azo dye that causes tumors in the liver or bladder of rats, mice and rabbits. It is considered a possible human carcinogen and mutagen, and is classified as category 3 carcinogen by the International Agency for Research on Cancer (IARC). In some countries, the harmful contaminant is usually added to help retain the typical reddish colour of pepper-based products and meat derivatives, but this particular application is not allowed in the European Union and many other countries [9], [10]. Based on the method published by European Commission (Corrected method for the detection of Sudan, NEWS notification: 03/99), Sudan I was extracted by acetonitrile and the extract was analyzed by HPLC in reversed phase (RP) after filtration. A variable UV/VIS detector is used for quantification, and as a result, the limit of quantization (LOQ) and limit of detection (LQD) were originally 0.100 and 0.030 μg/mL. After the application of SR, the LOQ and LOD were improved to 0.020 and 0.006 μg/mL, respectively. So SR can remarkably improve the signal-to-noise ratio (SNR) of HPLC-UV/VIS and make it possible to perform the trace analysis of Sudan I in some kinds of pepper-based products.

Section snippets

Theory and algorithm

Nonlinear Langevin equation has been applied to describe the phenomenon of SR. It is defined as follows [1], [11], [12]: $\frac{d x}{d t} = - U^{'} (x) + M I (t) + C ξ (t),$ where $I (t) = S (t) + N (t)$ denotes an input signal $S (t)$ and the intrinsic noise $N (t)$ ; $ξ (t)$ is the external noise, M and C are the adjustable parameters. $U (x)$ is the simplest double-well potential with the constants a and b characterizing the system $U (x) = - \frac{1}{2} a x^{2} + \frac{1}{4} b x^{4} .$

The symmetric double-well shows that the minima are located at $\pm x_{m}$ , where $x_{m} = {(a / b)}^{1 / 2}$ . A

Chromatographic and detection conditions

The HP1090 system was equipped with SPD-10A VP UV-VIS detector. The N2000 chromatography data system (Zhejiang University Star Instrument Technology Co. Ltd.) was used with sampling frequency of 10 Hz. The samples were separated on a Lichrospher C₁₈ ( $150 mm \times 4.6 mm$ ID, 5 μm) column. The mobile phase was a mixture of acetonitrile-acidified water (165 ml acetic acid plus 1000 ml water) ( $20 : 80$ , V/V) at a flow-rate of 1.0 mL/min. The wavelength was set at 478 nm.

Reagents

Sudan I (1-phenylazo-2-naphthol)

Optimization of system parameters a and b

In Eq. (2), the parameters a and b not only define the height of the potential barrier ( $Δ U = a^{2} / 4 b$ ), but affect the profile of the potential well. When the input signal was fixed, the parameters a and b affected the quality of final output signal directly. Therefore, it is necessary to optimize the parameters a and b in order to get a good output result. According to previous literature [1], taking into account the shape of the output signal, the quality of final output signal can be evaluated by

Conclusions

The quantitative analysis of Sudan I in some kinds of pepper-based products showed that SR could not only improve the detection limit and quantification limit but kept a good quantitative linearity between concentration and peak strength acquired by SR. The method presented in this Letter can enhance the sensitivity of instruments and make it possible to detect the weak signal accurately, which could not be detected effectively before. With the deep research of the theory of stochastic

References (15)

Z. Pan
Chemom. Intell. Lab. Syst.
(2003)
S. Fauve
Phys. Lett. A
(1983)
R. Zhu
Phys. Lett. A
(2001)
W. Guo
Chemom. Intell. Lab. Syst.
(2005)
X. Wu
Talanta
(2003)
W. Zhang
Anal. Chim. Acta
(2005)
W. Zhang
J. Chromatogr. B
(2006)

There are more references available in the full text version of this article.

Cited by (14)

Sweet and bitter tastants specific detection by the taste cell-based sensor
2012, Biosensors and Bioelectronics
Citation Excerpt :
It is demonstrated in numerous research works that the strength of weak signal will increase and that of intrinsic noise will decrease by the cooperation of proper additional noise, weak signal, and nonlinear system. SR has been applied in various scientific fields to improve analytical signals (Wu et al., 2003; Duan and Abbott, 2007; Leng et al., 2007; Ye et al., 2006; Ritaban et al., 2006; Sun and Lei, 2008; Ueda, 2010). Weak taste signals are amplified by SR method.
Sweet and bitter tastants specific detection by cell-based sensor is investigated in this paper. Human enteroendocrine NCI-H716 cells, expressing G protein-coupled receptors and sweet receptors (type 1, member 2/type 1, member 3), and human enteroendocrine STC-1 cells, expressing G protein-coupled receptors and bitter receptors (type 2 members) are used as sensing devices. The HEK-293 cells, without taste receptor expression, are used as negative control. The electrochemical impedance spectrum data is recorded and processed by bistable stochastic resonance for signal-to-noise ratio calculation. NCI-H716 cell-based sensor selectively responds to sweeteners and sweet tastant mixtures. STC-1 cell-based sensor selectively responds to bitter tastants and bitter tastant mixtures. The tastants species and concentrations can be decided by signal-to-noise ratio parameters. HEK-293 cell-based sensor lacks the tastants discriminating ability. The taste cell-based sensor is easy to prepare and operate. This work offers a useful way in gustatory mechanism research.
Molecularly imprinted microspheres as SPE sorbent for selective extraction of four Sudan dyes in catsup products
2011, Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences
Citation Excerpt :
However, this method suffers from the waste of organic solvents, is time-consuming and has potential toxicity hazards [13]. Other sample preparation methods, such as solid-phase extraction (SPE) [14,15], liquid phase microextraction [16], matrix solid-phase dispersion [17], dual solvent-stir bars microextraction [14], ultrasonic-assisted extraction [18] and centrifugal sedimentation [19,20] are also reported for Sudan dyes analysis. However, the selectivity of these pretreatment procedures needs to be further improved.
A highly selective molecularly imprinted solid-phase extraction (MISPE) coupled with high performance liquid chromatography (HPLC) ultraviolet–visible detection was developed for the simultaneous isolation and determination of four Sudan dyes (I, II, III and IV) in catsup products. The novel molecularly imprinted microspheres (MIM) were synthesized by aqueous suspension polymerization using phenylamine and naphthol as template, which showed high affinity to Sudan dyes in aqueous solution. In order to develop a selective extraction protocol for simultaneous determination the four Sudan dyes from catsup products, the molecular recognition properties of MIM as a SPE sorbent were evaluated. Under the optimized condition, good linearity was obtained from 0.01 to 2.5 μg g⁻¹ (r² ≥ 0.9990) with the relative standard deviations of less than 3.4%. This proposed MISPE-HPLC procedure eliminated the effect of template leakage on quantitative analysis and could be applied to direct determination of four Sudan dyes in complicated food samples.
A review of analytical techniques for determination of Sudan I-IV dyes in food matrixes
2010, Journal of Chromatography A
Sudan dyes are a family of lipophilic azo dyes, extensively used in industrial and scientific applications but banned for use as food colorants due to their carcinogenicity. Due to the continuing illicit use of Sudan dyes as food colorants their determination in different food matrices – especially in different chilli and tomato sauces and related products – has during the recent years received increasing attention all over the world. This paper critically reviews the published determination methods of Sudan I–IV dyes. LC–UV–vis and LC–MS are the dominating methods for analysis of Sudan I–IV dyes. Sudan dyes are usually found in food at mg kg⁻¹ levels at which it may be necessary to use a preconcentration step in order to attain the desired detection limits. Liquid–solid extraction is the dominating sample preparation procedure. In recent years it has been supplemented by ultrasonic-assisted extraction and pressurized liquid extraction. Various solid phase extraction types have been used for sample cleanup. The large majority of works use conventional C18 columns and conventional LC eluents. Traditionally the UV–vis absorbance detection has been the most frequently used. In the recent years MS detection is applied more and more often as it offers more reliable identification possibilities.
Comparison of dual solvent-stir bars microextraction and U-shaped hollow fiber-liquid phase microextraction for the analysis of Sudan dyes in food samples by high-performance liquid chromatography-ultraviolet/mass spectrometry
2008, Journal of Chromatography A
Two sample preparation methods, dual solvent-stir bars microextraction (DSSBME) and U-shaped hollow fiber–liquid phase microextraction (U-shaped HF–LPME), are proposed and critically compared for high-performance liquid chromatography (HPLC)–ultraviolet (UV)/mass spectrometry (MS) analysis of Sudan dyes in this paper. In DSSBME, the organic solvent was confined to a pair of hollow fiber membrane fixed on a stir bar, which can stir by itself, while the hollow fiber in U-shaped HF–LPME was fixed by two microsyringes. The significant factors affecting the microextraction of Sudan dyes in both microextraction techniques have been examined and no obvious difference in the effect of extraction solvent, pH and salt concentration on the extraction efficiency of Sudan dyes was observed except extraction time and stirring speed. Both microextraction techniques were similar in terms of analytical performance from aqueous solutions (LODs ranged from 0.09 to 0.95 μg L⁻¹ by HPLC–UV and 2.5–6.2 μg L⁻¹ by HPLC–MS; the absolute LODs ranged from 0.9 to 11.25 pg by HPLC–UV and 5–21.2 pg by HPLC–MS), however, DSSBME was more stable (lower stirring speed required), less sample consuming and much shorter time required to reach extraction equilibrium; while U-shaped HF–LPME was easier to operate and no more special device required. The two microextraction techniques combined with HPLC–UV/MS were successfully applied to the analysis of real samples including strawberry sauce, capsicum oil, salted egg, and two kinds of chilli sauce. Although the LODs of HPLC–UV are lower than that of HPLC–MS by a factor of 10 in this work, the absolute LODs for both HPLC–UV and HPLC–MS are comparable. HPLC–UV cannot identify the suspicious peaks at the same retention time as that of Sudan II and III in salted egg, while HPLC–MS can give exact information of Sudan I–IV in real sample analysis and is more reliable. The sensitivity of HPLC–MS is enough for real sample analysis.
Improved stochastic resonance algorithm for enhancement of signal-to-noise ratio of high-performance liquid chromatographic signal
2007, Analytica Chimica Acta
Citation Excerpt :
The algorithm based on the theory of stochastic resonance (SRA) can be used to determine weak analytical signals quantitatively, and has been applied successfully to the analysis of the Raman spectra from tetrachloromethane (CCl4) that were suffered from high-level noise and to the detection of weak chromatographic signals from carbon dioxide (CO2) [3]. In our previous works, SRA was used to determine the plasma concentrations of several drugs [9–11], to determine Sudan I in pepper-based products [12], and to determine chloramphenicol residues in milk [13]. SR is a powerful approach to amplify sub-threshold weak signals, but the optimization of nonlinear system parameters is the difficult problem always obsessing the analysts while stochastic resonance is applied to detection of weak signals.
Based on the theory of stochastic resonance, an improved stochastic resonance algorithm with a new criterion for optimizing system parameters to enhance signal-to-noise ratio (SNR) of HPLC/UV chromatographic signal for trace analysis was presented in this study. Compared with the conventional criterion in stochastic resonance, the proposed one can ensure satisfactory SNR as well as good peak shape of chromatographic peak in output signal. Application of the criterion to experimental weak signals of HPLC/UV was investigated and the results showed an excellent quantitative relationship between different concentrations and responses.
Fast Simultaneous Determination of Eight Sudan Dyes in Chili Oil by Ultra-High-Performance Supercritical Fluid Chromatography
2019, Journal of Analytical Methods in Chemistry

View all citing articles on Scopus

View full text

Stochastic resonance is applied to quantitative analysis for weak chromatographic signal of Sudan I

Abstract

Introduction

Section snippets

Theory and algorithm

Chromatographic and detection conditions

Reagents

Optimization of system parameters a and b

Conclusions

Chemom. Intell. Lab. Syst.

Phys. Lett. A

Phys. Lett. A

Chemom. Intell. Lab. Syst.

Talanta

Anal. Chim. Acta

J. Chromatogr. B